The long-term goal of this research is to benefit the development of communication aids for people with dysarthria and their conversation partners by increasing our knowledge of speech perception for disordered speech. Communication entails mutual speaker-listener interactions (Lindbloom 1990), thus it is critical to consider both speaker-oriented factors and listener-oriented factors that lead to intelligibility enhancement for speakers with dysarthria. Compared to speaker-oriented factors such as speaking style manipulations, less attention has been directed to listener-oriented factors such as signal-dependent factors (e.g., familiarity to the speech) and signal-independent factors (e.g., topic cue). Concerning signal-dependent factors for listeners, previous studies generally suggest that experience with dysarthric speech improves a listener's ability to perceive the speech. However, researchers do not have a full understanding of what elements of familiarization training restrict or promote familiarization effects. Further, a question remains largely unanswered concerning the underlying mechanism of intelligibility improvement due to a listener's familiarity with the speech. A better understanding of this mechanism will reveal methods to expedite the familiarization process, thereby allowing speakers with dysarthria to interact with new conversation partners with less pre-communication overhead. The perception theory hypothesizes that when speech is degraded, perceptual reorganization is necessary for listeners to understand the speech. In addition, it is hypothesized that listeners' perceptual reorganization leads to intelligibility enhancement due to improved mapping of acoustic features of speech signal into phonological representations. The proposed research will advance the perceptual reorganization hypothesis, by comparing the effects of different familiarization paradigms on the word and phoneme intelligibility of dysarthria and by seeking to understand the acoustic basis of familiarization effects. The target speech will be a subset of the Universal Access data we collected. It offers unique advantages for the proposed studies: it contains recordings of a large number of words produced by speakers with CP-related dysarthria in a wide range of severity as well as age/gender-matched control speakers. Results of this research are critical for advancing our knowledge of speech perception in general, as well as for providing empirical evidence for listener-oriented intervention techniques in the management of dysarthria. Findings will also directly guide the development of communication aids such as voice-input voice-output communication aids in the next phase of this work.